GCS Assist is the multi-award-winning generative AI tool built specifically for UK Government Communications.

Providing a detailed analysis that identifies significant as well as subtle issues in content that may not align with our government's GBA+ policy.

Analyzed multiple spreadsheets with web analytics data for a set date range to provide a detailed analysis of the data.

Using Copilot AI to provide a wide and deep analysis of legislation vs. web content in a brief report.

The Latent Outbreak Visibility System is an intelligence and epidemiological tool that brings together publicly available data and evidence from across the many different actors responding to the Ebola outbreak

How we built a tool that gives colleagues from across DBT a way to ask questions in plain English and get clear, trusted answers drawn entirely from intranet content.

When I.AI closed their Redbox product in 2025, BIST had already diverged substantially from the original platform. Two years of user research and iteration created something distinct, our suite of tools called DBT Assist.

The Global Supply Chains Intelligence Programme (GSCIP) team has won the Analysis in Government Impact Award for informing major policy decisions, improving operational outcomes and driving analytical and AI innovation while saving government millions.

The DBT Assist team shares how they safely connected DBT’s AI tools to live customer relationship management data, overcoming a common challenge faced across government.

How Spain's innovation agency tested AI tools to improve the assessment of social impact in start-up grant applications.

AI tools across chat, email, and voice, all drawing on the agency's own website as their single source, to help close the gap between the volume of public queries and the capacity to answer it.

An exploratory study at a Malaysian hospital evaluates a hybrid AI system for drug monitoring, proposing four key safeguards for safer LLM deployment in high-risk clinical settings.

The Integrated Value Network (IVN) uses artificial intelligence and graph-based knowledge management to create a connected view of governance.

IPC developed a targeted, internal Artificial Intelligence (AI) tool that allows staff to search the office's published orders and decisions using plain language.

Using a large language model to organise large volumes of open-ended public service feedback into themes, sentiments, and points of friction.

A hub of course-scoped Socratic AI tutors, paired with curated visual materials, giving healthcare students on-demand support in high-failure subjects.

Have you talked to a tree today? The Talking Treebot gives Berlin's city trees a voice with the help of artificial intelligence.

The first application to emerge from the Qtrees research project is Baumblick. Baumblick is a web application that gives an insight into and an outlook on the condition of our city trees in Berlin.

Public administration processes large amounts of information on a daily basis, but does not have a digital knowledge management system for the most part. With our prototype Parla, we are testing the possibilities of using artificial intelligence to support this process of information retrieval.

A new AI assistant for the Berlin state administration that supports administrative staff in their daily work with smart AI features and makes routine tasks more efficient.

Kansas City, Missouri, became the first city to automate FEMA preliminary damage assessments with Bellwether at X, the Moonshot Factory, reducing the process from days to minutes with a novel AI solution.

Collects and categorises symbols, with automated analysis for law enforcement and community workers.

Balances operational needs with staff skills, seniority, and preferences.

Draws on historical data to flag risks early for intervention.

Pulls together fire, transit, and crisis data to flag higher-risk locations.

Uses image detection to support frontline staff in the field.

Forecasts high-risk locations and tracks how interventions land, using previously siloed partner data.

A bilingual voice AI mentor, piloted in a state skills programme, that guides engineering students through building AI projects by asking questions rather than giving answers.